Membranes have been made that are hygro-responsive — their wetting properties change when immersed in water. This striking property allows the membrane to separate emulsions into their oil and water constituents.
Oil and water don't mix, so the saying goes — unless they form an emulsion, in which case it is difficult to get them apart. Reporting in Nature Communications, Tuteja and colleagues1 describe a simple, scalable method of great potential for separating such 'oily water' mixtures. They have developed membranes whose surfaces are extremely repellent to oil, but which allow water to permeate freely when oily water is filtered through them, so that the retained liquid is principally oil. Unlike mechanical systems such as centrifuges or settling tanks, which separate oil from water only if the oil phase is a distinct dispersion of droplets, the authors' 'smart' membranes separate emulsions highly efficiently. Such hygro-responsive membranes could be developed to clean up oil-contaminated sea water.
Tuteja and colleagues previously reported2,3 superoleophobic surfaces — ones that resist wetting by liquids that have extremely low surface tension, such as oils and alcohols. The key to making them was the recognition that the surfaces' texture is crucial for superoleophobicity. In particular, re-entrant surface curvature (surfaces that have concave topographic features) is required2. So, by making surfaces that have an appropriate chemical composition, roughened texture and re-entrant surface curvature, the authors prepared materials that were extremely resistant to wetting by several liquids. These surfaces can be thought of as omniphobic, because they are highly repellent to water as well as to oils.
More recently, Tuteja's group went further by developing oleophobic membranes4 that separate oily water emulsions when an electric field is applied across the membrane. This enabled 'on-demand' separation of millilitres of emulsion, but it is questionable whether the system could be used at an industrial scale. Although electrically enhanced processes5 were an active research area in the 1980s and 1990s, commercial developments have not followed because scaling up is a problem.
The membranes now reported by Tuteja and colleagues1 are different. The authors describe them as hygro-responsive, a word that derives from the Greek hygros, which means wet. This description is certainly pertinent, because wetting of the membranes by water — along with wicking and capillary flow — is vital for their separation properties. The authors prepared their membranes by coating either a stainless-steel mesh or a polyester fabric with a blend of a polymer and an oligomeric material. The resulting non-wetted membranes are both superoleophobic and hydrophobic, but when they are wetted, molecules at the surface of the coating reconfigure in such a way as to enable excellent water permeability while retaining superoleophobicity. This reconfiguration could be attained within a few minutes, which means that the time taken to 'activate' a membrane with water will not be a problem in industrial applications.
A similar reconfiguration has been observed at the surfaces of other polymer films, such as poly(methyl methacrylate), for which the relationship between molecular surface rearrangement and wettability has been well characterized6. It has also been noted7 that surfaces that have been chemically modified by the attachment of amphiphilic macromolecules (polymers that have both hydrophilic and hydrophobic properties) can lead to 'switchable wetting', in which the surface's wetting properties change depending on the properties of fluids to which they are exposed. By taking these materials through several wetting and drying cycles with water, it was shown that surface reconfiguration in these systems is reversible.
Two aspects of the hygro-responsive membranes1 are particularly striking. First, the water flux through the steel-mesh membrane is exceptionally high at around 43,000 litres per square metre per hour (more than 10 litres per square metre per second). This is more than 1,000 times that of a typical industrial ultrafiltration membrane unit. Second, the fact that the authors' technique for making hygro-responsive membranes can be applied to textiles and other surfaces is exciting, because this will enable a range of options to be explored, parallelling the wide range of module types in the membrane industry. Filtration modules based on hygro-responsive membranes, and capable of treating many tonnes of oily water each day, may well emerge soon.
The authors describe their separation technique as a capillary-force-based separation method — that is, one that exploits the difference in capillary forces acting on the individual phases of oily water as it interacts with the membrane. This is a fair description of the process. More questionable is their statement1 that their process is “solely gravity driven”. Although gravity can certainly be used to bring oily water emulsions into contact with the membrane, if an emulsion was pumped between two hygro-responsive membranes, I am confident that water would penetrate through both membranes irrespective of their orientation (and therefore of the influence of gravity). A simple experiment could be performed to test this.
Tuteja and colleagues also provide an equation for the breakthrough pressure of their membranes — the maximum pressure difference across the membrane at which the material prevents the permeation of oil. This enables pumped systems to be designed that use the membranes to separate oil–water emulsions. Such systems would be low-pressure systems in the eyes of process engineers, and would therefore have low operating costs. A design for one possible system is shown in Figure 1.
The authors separated emulsions of water and rapeseed oil as proof of concept of their work. In a related study8, others have separated mixtures of water and hexadecane (a diesel-like hydrocarbon). However, in the real world, filtration processes suffer from fouling and biofouling of the membranes. Further work using sea water and oil, and a systematic study of possible foulants, should therefore be undertaken to assess the commercial potential of these exciting new membranes.
Kota, A. K., Kwon, G., Choi, W., Mabry, J. M. & Tuteja, A. Nature Commun. 3, 1025 (2012).
Tuteja, A. et al. Science 318, 1618–1622 (2007).
Tuteja, A., Choi, W., Mabry, J. M., McKinley, G. H. & Cohen, R. E. Proc. Natl Acad. Sci. USA 105, 18200–18205 (2008).
Kwon, G. et al. Adv. Mater. 24, 3666–3671 (2012).
Bowen, W. R. in Membranes in Bioprocessing (eds Howell, J. A., Sanchez, V. & Field, R. W.) Ch. 8 (Blackie, Chapman & Hall, 1993).
Horinouchi, A., Atarashi, H., Fujii, Y. & Tanaka, K. Macromolecules 45, 4638–4642 (2012).
Howarter, J. A., Genson, K. L. & Youngblood, J. P. ACS Appl. Mater. Interfaces 3, 2022–2030 (2011).
Howarter, J. A. & Youngblood, J. P. J. Colloid Interface Sci. 329, 127–132 (2009)
About this article
IOP Conference Series: Earth and Environmental Science (2020)
Journal of Macromolecular Science, Part A (2020)
Unraveling the Effect of PEG Chain Length on the Physical Properties and Toxicant Removal Capacities of Cross-Linked Network Synthesized by Thiol–Norbornene Photoclick Chemistry
ACS Omega (2020)